A finance buyer almost never searches for what is physically in the frame. They search for the idea the frame stands in for: saving, borrowing, growth, a decision about money. Keywording this category well comes down to two disciplines pulling against each other, naming concepts that are genuinely true of the image while stripping out the brands these frames are full of.
What a finance buyer types, and why it is rarely in the frame
Most stock categories reward literal description. A photograph of a loaf of bread is found by someone who typed bread. Finance breaks that symmetry, because the people licensing these images are illustrating abstractions. A bank's landing page needs mortgage. A fintech explainer needs onboarding. A personal finance article needs emergency fund. None of those words name an object, and the objects available to a photographer in this category are a short, repetitive list: cards, screens, notes, coins, handshakes, pens over paperwork.
That mismatch is why a tagger that reads the frame accurately can still produce a keyword set that earns nothing. Every term is true and every term is useless, because the buyer's search never contained any of them. The work is translation, not description.
The concept has to be true of the picture
The obvious failure mode of conceptual keywording is stuffing every abstraction a finance editor might want onto every file. Agencies close that door explicitly. Adobe's guidance on titles and keywords tells contributors to use only relevant keywords to avoid content refusal, and Shutterstock treats metadata unrelated to the depicted subject as grounds for rejection under its commercial metadata requirements.
So the test for a concept keyword is not whether a buyer might search it. It is whether a reviewer looking at the image would accept that the image depicts it. A person frowning at a laptop and a stack of envelopes can carry debt and financial stress; it cannot carry retirement planning, because nothing in the frame says retirement. One or two concepts that survive that test beat eight that do not, and a refused file earns nothing at all.
Adobe also says keyword order is essential and should reflect the content's importance, and asks contributors to place the most important and relevant keywords in the first ten positions because they have the greatest influence on search ranking. In finance that makes the ordering decision a real one: the concept usually deserves a place ahead of the object, because the concept is what was searched.
Brand density is this category's defining constraint
No other commercial niche puts this many trademarks in front of the lens. Payment cards carry scheme marks. Bank branches carry signage. Laptops and phones carry maker logos, and their screens carry software interfaces that are themselves protected work. A finance set shot without planning will be substantially unusable commercially, and the metadata is where contributors most often compound the problem rather than contain it.
Both major agencies state the metadata rule in almost the same words. Adobe tells contributors to keep trademarks, brand names and personal information out of the submission. Shutterstock says descriptions and keywords for commercial content cannot include trademarks such as brand names or product names. There is no nuance to work with here and no partial credit: a brand name in the keywords is a refusal route, not a discoverability tactic.
What trips people up is the inverse move. A contributor who has carefully retouched a card mark out of the frame then writes the scheme name into the keywords, reasoning that buyers search it. That inverts the whole point. The clean frame is the asset, and the honest keyword for it is unbranded, not the brand that is no longer there.
A visible logo is also usually a shooting or retouching problem rather than a tagging one. Metadata cannot license a mark that is legible in the image. The decision tree runs the other way: if the mark is readable, either it comes out in post, or the file goes editorial, or it does not go up at all. Only once that is settled is there a keyword set worth writing.
The same payment frame, keyworded two ways
A card-and-laptop payment frame as a vision model tags it, against the same frame keyworded for finance buyers, with the reason each term earns or wastes a slot. The trademark rules quoted come from Adobe's titles and keywords guidance and Shutterstock's contextual metadata standards.
Numbers on a screen are their own metadata problem
Shutterstock's metadata standards include a line that reads like boilerplate until you shoot this category: do not include sensitive personal information, such as social security or financial numbers. In most niches that rule never comes up. In finance it is a live constraint, because the frames are full of account numbers, card digits, balances and reference codes, and the instinct when writing a description is to describe what the screen says.
The safe habit is to keep every number in the metadata generic. A balance is a balance, not the figure shown. An account is an account, not a string of digits. The same caution applies in the frame itself: real card numbers, sort codes and account references belong nowhere in a commercial file, and a mocked-up interface with obviously invented values is both safer and more licensable than a real one blurred badly.
Currency, country and the editorial line
Money is the one finance subject where specificity helps rather than hurts. A buyer illustrating a story about a particular economy needs that economy's notes, and the generic keyword money cannot deliver it. Naming the currency and the country is legitimate, specific and searched, and it is one of the few places in this category where a literal term outperforms a conceptual one.
Location metadata rewards the same precision. Shutterstock treats the location field as optional but asks that it be as specific as possible when supplied, naming neighbourhood, town or city as the useful level rather than a region or a continent. For a financial district frame that is straightforward value: the district name is often exactly what was typed.
Where currency and recognisable institutions push a file toward editorial, the caption rules change rather than relax. Shutterstock's contextual metadata standards require documentary editorial content to carry the day, month and year the content was taken plus the geographic location, and a missing or inaccurate date or location is itself an editorial metadata rejection. A branded bank frontage can work as editorial with a dated, located, factual caption; the same frame with a promotional caption works nowhere.
Where the generic finance nouns waste your slots
Finance has a standard set of words that feel productive and compete with hundreds of thousands of near-identical files: money, business, finance, success, growth, investment, wealth. They are rarely wrong and almost never decisive. A file that ranks nowhere on finance might rank well on invoice reconciliation or late payment reminder, because the second pair describes a situation rather than a sector.
The filler adjectives are worse, because they are not even true in a checkable sense. Adobe asks for only relevant keywords and tells contributors to use each keyword once, which rules out both the decorative adjective and the habit of padding a set with singular and plural forms of the same term. Every slot spent on stunning or professional is a slot not spent on the situation the image actually shows.
Working inside the counts the agencies enforce
The published limits leave less room than the ceilings suggest. Adobe accepts up to 49 keywords per submission and asks for titles kept short, ideally under 70 characters. Shutterstock requires a minimum of 7 keywords and allows a maximum of 50, with the description a factual sentence capped at 2048 characters. Both require English, with narrow exceptions for scientific names, place names and foreign terms already common in English.
In practice a finance file rarely supports 49 defensible keywords. Once the brands are out, the filler is out, and every concept has passed the would-a-reviewer-agree test, a careful set lands somewhere in the twenties. Filling the remaining slots to reach the ceiling is how an otherwise clean file acquires an irrelevant-metadata rejection.
Keywording a finance batch in six passes
A pass-by-pass order for a finance shoot that settles the commercial or editorial split and the brand removals before any keyword is written, then names the transaction, the concept and the usability terms, and finally checks the set against Adobe's 49-keyword ceiling and Shutterstock's 7-to-50 range.
Which tools actually help on finance files
Vision models are good at the object layer and blind to the two layers that decide a finance file. They will name the card, the laptop and the hand reliably, and they will not reliably tell you whether a mark is legible or whether the concept you want is defensible. That makes them a drafting aid here rather than an answer. A general metadata editor such as Bridge or Lightroom has the opposite profile: full control, no judgement, and no knowledge of any agency's rules, which is a trade-off worth understanding before committing a workflow to it, and the comparison between the two sets out where each one stops.
What is worth paying for in this category is policy awareness: a banned-term list that catches brand names before submission rather than after a rejection, and keyword sets you can review in bulk instead of file by file. The six-tool comparison for finance content goes through the options in detail. If the frames are generated rather than photographed, the agency rules for AI-generated images add a disclosure layer on top of everything above.
When to use which
For a handful of finance files a month, keyword by hand against the agency's own published rules and skip the tooling question entirely; the judgement calls dominate and nothing automates them. For a regular finance output, generate a draft set, then spend the saved time on the brand sweep and the concept test rather than on typing, because those are the two passes that decide whether the file is licensable at all.
Rastock AI sits in the second case: metadata generation with per-agency rule filtering, banned and mandatory keyword handling, and delivery to 10+ agencies over FTP and SFTP, with IPTC and CSV export if you would rather keep the metadata and leave. The feature list covers what it does and does not decide for you, and the pricing page includes a 14-day free trial with no card required. It will not tell you whether a logo is legible; no tool will, and on finance files that remains the photographer's call.
Frequently asked questions
How many keywords should a finance stock photo have?
Adobe accepts up to 49 keywords per submission and Shutterstock requires a minimum of 7 and a maximum of 50. A finance file rarely supports the ceiling honestly. Once brand names, filler adjectives and undefendable concepts are removed, a careful set usually lands in the twenties, and padding beyond that invites an irrelevant-metadata rejection.
Can I name a card scheme or bank in the keywords?
No. Adobe asks contributors to keep trademarks, brand names and personal information out of the submission, and Shutterstock states that descriptions and keywords for commercial content cannot include trademarks such as brand names or product names. If the mark is legible in the frame, that is a retouching or editorial decision rather than something metadata can resolve.
Should I describe the numbers shown on a screen?
Keep them generic. Shutterstock's metadata standards say not to include sensitive personal information such as social security or financial numbers, which matters in this category more than almost any other. Describe a balance as a balance rather than quoting the figure, and keep real account references, card digits and sort codes out of both the frame and the metadata.
Are conceptual keywords like growth or success worth adding?
Only where a reviewer looking at the image would agree it depicts them. Adobe tells contributors to use only relevant keywords to avoid content refusal, and Shutterstock treats metadata unrelated to the depicted subject as grounds for rejection. Broad sector words also compete with enormous numbers of near-identical files, so a specific situation usually outperforms an abstraction.
Should I name the currency in a photo of banknotes?
Yes, and the country with it. Currency is the one finance subject where literal specificity beats conceptual keywording, because a buyer illustrating a particular economy needs that economy's notes and the generic word money cannot deliver it. Naming the currency is specific, true of the frame and actually searched.
When does a finance photo have to go editorial?
Usually when a recognisable institution or a legible brand mark cannot be removed. Editorial then brings its own metadata requirements: Shutterstock requires documentary editorial content to include the day, month and year the content was taken plus the geographic location, and a missing or inaccurate date or location is itself an editorial metadata rejection.